Previous chapter

VCR_LTER_III

RESULTS OF PRIOR NSF SUPPORT

Long-Term Ecological Research on Landscape Development and Ecological Processes in a Tide-Dominated Barrier-Lagoon Complex: LTER II

NSF DEB 92-11772; $1,241,406; September 1, 1992 to October 31, 1994.

INTRODUCTION

VCR LTER II focused on ecosystem change: disturbance, succession, and ecosystem state change. We addressed the following hypotheses: 1) The availability of fresh water, the salinity of available water, the frequency of tidal inundation, and the dynamics of sediment deposition control succession at the VCR; 2) Gradients in salinity, tidal inundation, organic matter accumulation, primary productivity, and nutrients, are functions of the land slope, sea level, and astronomical and storm tides; and 3) Ecological state changes are triggered by systematic long-term phenomena such as sea level rise and sedimentation, as well as by intense disturbances such as coastal storms.

Our studies are conducted at the Virginia Coast Reserve (VCR) on Virginia's Eastern Shore (Fig. 1). The Nature Conservancy is the steward of our site. During the two years of LTER II, we extended our research from LTER I, continued our systematic long-term measurement program, initiated three long-term experiments designed to bring about transitions in ecosystem states, incorporated new technologies needed to obtain high resolution measurement of critical landscape attributes (sea, land, and water table levels), and further developed our research infrastructure in support of long-term research.

From January 1992 through the present, 31 journal articles have appeared in print,14 are in press (1994 citations) and 13 are under review (1994 citations). VCR PIs have contributed 15 book chapters in print and 4 in press and 2 books which offer synthesis of research and theory at the site and also support intersite and network initiatives. Twelve theses and dissertations were completed.

This document summarizes our scientific activities for each of the 4 main research sites at the VCR: 1) The VCR Megasite; 2) The North Hog Chronosequence Site; 3) The Mainland Marsh Site; and 4) The Hog Island Bay Lagoons and Marshes Site. In addition, we review the progress of our long-term experiments; detail our accomplishments resulting from intersite, network, modeling and synthetic studies; and identify the relationship between past and proposed work.

THE VCR MEGASITE

Large-scale, long-term research at the VCR focuses on the causes and effects of ecosystem disturbance and change. For example, we detailed the hydrodynamics of the greatest coastal storm since 1942 (Davis and Dolan 1992); we derived estimates of the wave energies of 1447 coastal storms since 1942 (Dolan and Davis 1993) and analyzed the impact of these storms on island dynamics (Fenster and Dolan 1994). We calculated the occurrence probabilities of 100 barrier island plant species across the historical storm disturbance gradient (Fahrig et al 1993) (Fig. 5). We have also documented, using tree rings, individual tree response to hurricanes (Johnson & Young 1992). On longer time scales Foyle (1994), Foyle and Oertel (1994), Oertel and Kraft (1993) and Oertel et al. (1993) charted the drainage patterns of VCR lagoons and showed their continuity with paleochannels on the continental shelf. These finding stimulated a new model for the origin of the lagoon and marshes of mid-Atlantic barrier islands (Oertel and Kraft 1993).

Recent efforts include increased use of remotely sensed data as a tool for documenting ecosystem processes. We completed the GIS framework for the annual VCR colonial bird nesting survey and have included the 1993 GPS geo-referenced bird survey is in this system. In the long-term we will track changes in utilization of the landscape by colonial birds. In addition, we have mapped land cover and changes in land cover for the VCR using TM satellite imagery. We installed a NASA sun photometer to better correct satellite imagery for atmospheric effects; completed our first study of spectral reflectance of barrier island plants (Carter and Young 1993); and demonstrated the utility of TM, Spot and aerial photography for analysis for vegetation change (Porter and Callahan 1994).

Habitat composition and utilization at this large spatial scale remain strong research areas. Biogeographic surveys of island vegetation, lagoonal fish (Yozzo 1994), marcoalgae (Monti 1993), formainifera (Woo 1992), and insular small mammal populations (Porter and Dueser 1994, Scott and Dueser 1992, Halama and Dueser 1992) were completed in late 1993. Tissue samples were collected for allozyme analysis of small mammals on 11 islands; analysis of blood samples to test for the presence of Hantavirus were negative (Moncrief, pers. comm.). Forys and Dueser (1993) documented movements of rice rats (Oryzomys palustris) between Crescent and Parramore Islands, all of which originated on the smaller island. We found that rats dispersing between islands comprise a statistically random subset of the resident population with respect to age and sex, unlike overland dispersers which are predominantly juveniles. Our biogeographic surveys are conducted on a campaign basis, with expected surveys at intervals of 3-5 years.

THE NORTH HOG CHRONOSEQUENCE SITE

Harris (1992) identified and dated a chronosequence of dune and swale landscape elements between the ocean and the lagoon behind Hog Island. The youngest features are located near the active beach, and the oldest (dating from ~1871) are adjacent to the lagoon-side marsh (Fig. 2A). Several studies indicate that dune age is an important control on ecological development. For example, soil nitrogen (ammonium and nitrate), net N-mineralization, and total N-mineralization rates are all highest in the oldest terrain and lowest in the newest landscapes (Young et al 1992) (Fig. 2B). Decomposition rates are greater in older landscapes than in newer terrain, and total below ground biomass increases from 582 g/m2 to 3035 g/m2 with increasing age of swales through the 120 year chronosequence. Above ground and below ground production on the dunes decreases with increasing site age (Fig. 2B); available nitrogen increases with age in the chronosequence while nitrogen mineralization is not well explained by landscape age (Fig. 2B). Continued chronosequence research will focus on long-term responses of primary production, organic turnover, and nutrient cycling rates to changes in island hydrology, nutrient availability, herbivory and storm events.

Through extensive surveys we have established the relationship between perched freshwater reserves and local land elevation across scales from microtopographic variations on marsh surfaces to the scale of dune/swale systems. In this framework, coastal storms transport sand and elevate land surface levels, thereby permitting higher fresh water-table elevations (Stasavich and Hmieleski 1993). Interestingly, over 90% of the beach sand washed inland by storm events remains in place on the island; only 10% of the sand returns to the beach by winds (Cohn 1993; Clark 1993). In areas of Hog Island where storm overwash disturbance is frequent, clonal growth forms dominate (Fahrig et al. 1993). Both ocean side flooding due to storm surge and lagoon side flooding due to wind tides result in salinity stresses on terrestrial island vegetation. Young and co-workers have studied the effects of salinity on barrier island flora. Young et al. (1994) established the salitinty tolerances of individual plant species. Johnson & Young (1992) determined the sensitivity of pines to salt water flooding and the effects of pore-water salinity on germination (Sande & Young 1992) and primary productivity (Young 1992). These studies highlight the sensitivity of the VCR ecosystem to subtle changes in the relative elevations of the land surface and the fresh/salt water interface.

THE MAINLAND MARSH SITE

Our studies of marsh dynamics at the landscape level have lead us to propose a conceptual model of sea-level induced transition from terrestrial forest to estuary (Brinson et al. 1994). Important components of this model include self-maintaining properties and state changes due to disturbance or exposure to acute stress. The mechanisms responsible for maintenance and transition along the gradient from lagoon to upland are unique for each zone within the marsh. For example, the position of the upland-high marsh transition depends upon the slope of the marsh surface and the distance from a tidal creek. Along steep marsh surface slopes, elevation is the primary factor influencing the position of the transition, while along very gentle slopes distance from the nearest tidal creek is most important.

Along the gradient from low to high marsh at the mainland marsh site (Fig. 3A), root decay processes apparently are not affected by sediment physico-chemical differences, although root production is much greater in the mid-marsh zone perhaps due to tidal inundation conditions. Greater root production and slower root turnover occur in the mid-marsh sediments with highly variable salinities, high oxidation-reduction potentials, lower sediment saturation, and high sulfide concentrations. Differences in organic matter accumulation between high and low marsh areas are thus due to differences in root production rather than root decomposition (Blum 1993). Although position in the marsh and sediment pore water chemistry has little effect on root decay, Juncus roots decay two times faster than Spartina roots. This difference is consistent with the difference in the starting C:N ratios of 37:1 and 47:1 for Juncus and Spartina, respectively. We are comparing two methods for measuring below ground decay: (1) decay in undisturbed plots in which new organic matter production is prevented by removing above ground plant material and pruning roots to exclude new root growth, and (2) decay in buried litter bags -- and conclude that regardless of the type of organic matter (i.e., Juncus or Spartina) or location in the marsh, the vast majority of below ground organic matter is old and recalcitrant (Christian et al. 1993) and/or the removal of new below ground root material restricts decomposition (Tirrell and Blum 1992).

Field measurements and GIS analysis of aerial photography were used to correlate landscape evolution to sedimentary processes at the Mainland Marsh Site. The marsh area has increased by 8% over 50 years, primarily because of upland encroachment (Kastler 1993) (Fig. 3B). Lead dating of the sediments indicate a sediment accretion rate of 2 mm per year (Kastler 1993), which is consistent with earlier measures by Barr (1989) and Oertel (1992). Analysis of the clay mineralogy of marsh sediments indicate that this marsh developed on top of upland soils similar to those found in neighboring agricultural and forested areas (Robinson 1993).

At the Mainland Marsh Site and in other VCR tidal creeks, the rate of carbon cycling by bacteria is much lower than in nearby creeks of Chesapeake Bay despite high concentrations of inorganic N and P (MacMillin et al. 1992). The high inorganic nutrient pools combined with the low levels of bacterial productivity indicate that bacterial production is not limited by N or P and that the amount of carbon moving through the bacterial loop is low relative to Chesapeake Bay tidal creeks on the Delmarva Peninsula. Comparative studies of VCR marshes and tidal creeks with those of Chesapeake Bay will continue through funding from NOAA Sea Grant program. The Phillips Creek area in the Mainland Marsh Site will continue to be a focal point for research into the processes that govern the hydrology, geochemistry and ecology of the marsh system.

LAGOONS AND MARSHES SITE

Textural and mineralogical analysis of inorganic sediments from throughout the VCR revealed that each physiographic area (i.e., mainland, marsh surface, channel bottom) has a distinct sedimentological signature (Oura 1993; Robinson 1993). Regions of elevated paleotopography are identified as drowned uplands through the presence of kaolinite and vermiculite (similar to mainland soils) sampled in meter-deep cores. Preliminary 210Pb dating of sediment cores suggests that sedimentation rates in the VCR system are 1.0-2.2 mm/yr (Kastler 1993); major storm horizons are easily identified within the shallow cores (Kastler 1993; Robinson 1993).

Sediment texture places important constraints on the productivity of individual marsh plants. For example, sandy creek-bank areas that are flushed by semi-diurnal tides are highly productive; in contrast, sandy areas that are not subject to twice daily inundation achieve hypersaline conditions and have low Spartina alterniflora productivity (Robinson 1993). In addition, Hussey and Odum (1992) found substantial differences in marsh evapotranspiration (ET) across salinity gradients; the highest ET and LAI are measured in marshes with low salinity, suggesting that above ground primary production is salinity dependent. Extractable ammonium and porewater ammonium and phosphate in the surface layers of young salt marshes (10-13 years) are nearly identical to those in marshes >100 years old, indicating a more rapid chemical maturation than previously determined (Osgood and Zieman 1993a,b; 1994).

LONG-TERM EXPERIMENTS

Three long-term experiments were installed during 1992 and 1993. These experiments are designed to modify the relative levels of the land, the fresh water table and the lagoon waters. In our marsh-surface lowering experiment, sections of the marshes were lowered 15 cm to mimic a sea level change in excess of upward accretion of the marsh surface. Results from the first year indicate that biomass, stem density and stem heights all increased as a result of the relative increase in lagoon elevations (Fig. 4A). A method of lowering the marsh surface 5 cm has been developed and will be used in additional marsh lowering sites. The second long-term experiment, installed in June 1993, involves a solar powered pump to lower the local fresh-water table by 1 to 3 mm d-1. During 1993, we did not alter the water table but pumped back into the research site and observed the natural variations in water table at mm d-1 resolution. We have determined that we can monitor changes in the water table resulting from semi-diurnal tides, spring-neap tidal cycles, daily evapotranspiration and rainfall input (Fig. 4B). Our third experiment is a sea level manipulation. We installed (August 1993) a system of retaining walls across a marsh ecotone and a solar powered sea water pumping system that will permit modest changes in the duration of tidal flooding similar to those that are occurring on longer time scales with the current rate of sea level rise.

INTERSITE, MODELING AND SYNTHETIC STUDIES

Our theoretical, synthesis and modeling papers focus on the ecosystem response to global climate change (O'Brien et al. 1992; Ray et al. 1992; Shugart 1992a, b, c; Shugart, Leemans, and Bonan 1992a, b, c; Shugart and Prentice 1992, Shugart and Smith 1992b; Shugart, Smith and Post 1992; Smith et al. 1992a, b, c; Shugart 1993a, b; Shugart and Soloman 1993, Smith and Shugart 1993; Soloman and Shugart 1993; Davis et al. 1994; Hayden 1994a, b, c; Shao et al. 1994a, b); succession in terrestrial and upland ecosystems (O'Brien et al. 1992; Shugart 1992; Smith et al. 1992a and b; Solomon and Shugart 1992; Urban et al. 1992a, b; Friend et al. 1993; Fahrig et al 1993 (Fig. 5) and 1994; Lauenroth et al. 1993; Shao et al. 1993, 1994a, b); and on the structure and dynamics of coastal and pelagic ecosystems (Ray and Hayden, 1992a, b; Oertel 1993a, b, 1994, Oertel et al. 1994a, b, c; Christian 1994). Some of these efforts directly involved work for another LTER site (O'Brien et al. 1992; Lauenroth et al. 1993) or involved authors from another site (Shugart, Leemans, and Bonan 1992a, b, c; Smith et al. 1992; Lauenroth et al. 1993). We have also participated directly in providing published overviews of integrating regional models of ecosystems with social systems (Hayden 1994a) and of long-term data needs for long-term studies and modeling efforts (Hayden 1994 b).

RELATION TO FUTURE RESEARCH

The work completed to date, within the context of prolonged scientific discussion among the VCR PIs, has prompted the development of a stronger theoretical foundation for our proposed research (see Section 1: Project Description, pages PD 5-6). This is the most significant accomplishment of the last year and a half as it has brought all PIs into a common research framework. Studies reported on in this section will continue in the 6 years of proposed work.

VCR RESEARCH INFRASTRUCTURE

Since 1992, we improved our research infrastructure at both the UVa laboratories and at the VCR site. We have improved our laboratories for remote sensing, GIS and modeling with new equipment and equipment upgrades. In addition, we added new capabilities in the area of data management and information exchange (see Section 2, page PD 27). Stephen Macko joined the LTER PI team and brings a significant capability in the area of organic geochemistry and isotope biogeochemistry (see pages APDX 3 - 4). New PI James Galloway brings to the VCR LTER his analytical laboratory for precipitation and water chemistry and his field wet and dry deposition station on Hog Island. The addition of Tanya Furman gains VCR access to her mineralogical and x-ray diffraction laboratory. Added PIs Patricia Wiberg, G. Carleton Ray, Jerry McCormick-Ray and Iris Anderson bring expertise in sediment dynamics, marine mammal ecology, invertebrate ecology and trace gas emissions. We also air conditioned our site laboratory and lodging areas, built a new outdoor, screened-in sample processing facility, purchased a mm-resolution laser surveying system, and completed our cm (x,y,z)-resolution, kinematic GPS surveying system.

HUMAN RESOURCE DEVELOPMENT

Between 1992 and the present we have graduated 11 students with advanced degrees (8 MS and 3 Ph.D.) who have received part or all of their support from the VCR LTER core grant. We currently have 13 MS students and 5 Ph.D. students conducting research at the site. Thirteen of the published papers reported on in the results of prior NSF support involved graduate students. In addition, we had 4 REU students during the 1992 field season and following semester, and 5 REU students in 1993. Three of the four 1992 students are now in Graduate School and the fourth is applying to Graduate School. All of the 1993 REU students have plans to enter graduate school. We supported 4 additional undergraduate students in 1993 and they also have plans for graduate school.

Next chapter